Understanding the full costs of scholarly publishing across national and regional systems remains a challenge. What information is available is generally limited and the best data is usually confidential and private. The argument for Open Research Information is that the benefits of sharing outweigh the risks. We sought to examine this by building a large scale model of costs and savings in the scholarly publishing system using public information. Combining bibliographic resources including OpenAlex and OpenAIRE with cost information at the level of APCs (OpenAPC, DOAJ, and open datasets of archived list prices) and agreements (from full text agreements listed in ESAC) we can estimate the overall profile of OA publishing costs and use these to model the development of future costs under various sets of assumptions. The cost model for a given institution, consortium or country starts with the overall output volume and the proportion of outputs in each Open Access category. Within each category the proportion of outputs with a corresponding author is estimated. This is then compared to the set of outputs that are known to have been covered within agreements. For each category of paid open access (hybrid and APC-gold) we estimate list price APCs. For those outputs within agreements we also calculate the average cost per output (separating hybrid and gold), taking into account ‘publish’ and ‘read’ parts of the agreement, where applicable. Using this cost model, we then extrapolate to generate estimated future prices per output for each category (APC-gold, hybrid, within and outside agreements). These price estimates are then used to calculate the overall cost of each category in various scenarios. We use baseline estimates for subscription costs, and estimate current expenditure on repositories and diamond OA publishing venues as a baseline for infrastructure contributions. Scenarios covered in the trajectory modelling can include increasing the number and coverage of R&P deals, paying APCs out of contract, switching to only full gold OA publishing deals, shifting focus to diamond OA and/or repository-based OA, as well as mixed scenarios. The modelling approach can be validated by comparing data from open research information sources with data supplied by consortia directly. Consortial data can also be used to further refine the model, as these often include data that are not publicly available. This includes data on articles published as part of publisher agreements, as well as additional financial information on publisher agreements (including on subscriptions) and investments in repositories, diamond OA publishing and (other) open infrastructures. In this presentation, we show how a national, regional and global model of publishing costs can be built using Open Research Information. Through two case studies, we demonstrate the value of information shared by library consortia to this modelling approach, and encourage library consortia to make information on publishing costs information openly available.
Neylon et al. (Thu,) studied this question.